A Multi‐Structural Radiomic Digital Phenotyping Framework for Deciphering Preclinical Evolution of Degenerative Meniscal Tears
Shengfa Li, Hao Liu, Yanbin Zhu, Hang Zhang, Qifeng Tao, Wenbo Zhao, Xin Miao, Shanshan Yang, Tianyu Chen, David Hunter, Changhai Ding, Ting Wang, Juan Wang, Wei ChenABSTRACT
Degenerative meniscus tears (DMTs) are pivotal in knee osteoarthritis progression, yet conventional MRI often fails to capture the subvisual alterations preceding macroscopic failure. We developed a multi‐structural digital risk stratification framework (MSC‐RM) that integrates standardized radiomic signatures from six knee structures (femur, tibia, cartilages, and menisci) derived from 3D double‐echo steady‐state MRI with baseline clinical risk profiles. Utilizing a longitudinal cohort ( n = 674), our framework decoded complex degeneration patterns, achieving superior 4‐year predictive performance (AUC 0.871). Survival analysis revealed that high‐risk individuals had a 6.12‐fold greater hazard of developing DMTs than the low‐risk group. Furthermore, a reader test with five resident surgeons demonstrated that MSC‐RM significantly enhanced diagnostic consistency (Kappa increased from 0.38 to 0.55), effectively bridging the expertise gap. By shifting the paradigm from reactive lesion detection to proactive risk management, this framework provides a scalable strategy for precision preventive medicine and establishes digital biomarkers as critical tools for optimizing musculoskeletal health in aging populations.